askill
review-code-architecture

review-code-architectureSafety 95Repository

Review code changes for architectural consistency, correctness risks, security/performance concerns, and verification gaps; produce a prioritized, actionable review report.

0 stars
1.2k downloads
Updated 2/20/2026

Package Files

Loading files...
SKILL.md

Review Code Architecture

Purpose

Provide a rigorous, structured review of new or modified code to ensure architectural consistency, maintainability, and correct system integration.

When to use

Use this skill when:

  • A feature or endpoint is implemented and needs review
  • A refactor changed module boundaries
  • Performance, security, or maintainability concerns were raised
  • You are preparing code for merge/release

Inputs

  • The set of changed files (diff/PR) and a short feature description
  • Any relevant non-functional requirements (performance, security, reliability)
  • Existing architectural conventions (layering, module boundaries, style guides)

Outputs

  • A prioritized review report:
    • critical issues (must fix)
    • important improvements (should fix)
    • optional suggestions (may fix)
  • Concrete recommendations with rationale
  • Verification actions (tests, checks) to confirm fixes

Review rubric

1. Architecture & boundaries

  • Are responsibilities separated (UI vs data, controller vs service vs repository)?
  • Are dependency directions clean (no forbidden imports)?
  • Are abstractions appropriate (not premature, not leaky)?

2. API & contracts

  • Are input/output shapes explicit and stable?
  • Are error responses consistent and predictable?
  • Are public module exports minimal and intentional?

3. Error handling

  • Are errors mapped consistently?
  • Are unknown failures logged/tracked?
  • Are retries/idempotency considered where needed?

4. Security & privacy

  • Are auth/permission checks correct and enforced in the right places?
  • Are secrets and PII protected (not logged, not exposed to client)?
  • Are unsafe dynamic executions avoided (e.g., eval, unsanitized HTML)?

5. Performance

  • Any obvious N+1 queries, unbounded lists, or heavy renders?
  • Is caching/invalidation strategy correct?
  • Are slow paths measured or at least bounded?

6. Testing

  • Are critical business rules tested?
  • Are new endpoints covered by integration tests when feasible?
  • Are tests deterministic and maintainable?

7. Maintainability

  • Naming clarity and module structure
  • Code duplication vs reuse
  • Comments/documentation for non-obvious decisions

Steps

  1. Read the change intent (what problem it solves).
  2. Review the diff top-down:
    • contracts and boundaries first
    • then implementation details
  3. Identify risks and failure modes.
  4. Produce a structured report:
    • use the review report template
    • include evidence pointers (file/line or behavior)
  5. Provide verification actions per critical issue.

Verification

  • All changed files have been reviewed
  • Critical issues are identified and prioritized
  • Recommendations include concrete, actionable steps
  • Security implications are assessed for auth/permission changes
  • Verification actions are defined for each critical issue
  • Review output is structured and reusable

Boundaries

  • MUST NOT approve code without reading and understanding the changes
  • MUST NOT skip security review for auth/permission changes
  • MUST NOT leave vague feedback without actionable suggestions
  • SHOULD NOT block on personal style preferences that are not established standards
  • SHOULD focus on correctness and risk before cosmetics

Included assets

  • Templates:
    • ./templates/review-checklist.md
    • ./templates/review-report.template.md
  • Examples: ./examples/ includes a sample review report.

Install

Download ZIP
Requires askill CLI v1.0+

AI Quality Score

92/100Analyzed 2/22/2026

High-quality, comprehensive code architecture review skill with clear structure, actionable steps, detailed 7-category rubric, and reusable templates. Excellent clarity with well-defined boundaries and safety guidelines. Slightly penalized for being in .ai internal folder but otherwise highly reusable and well-suited for general application.

95
95
85
90
92

Metadata

Licenseunknown
Version-
Updated2/20/2026
Publisherwillyu1007

Tags

apisecurity